Dopamine agonists that activate D2 and D3 receptors (D2R, D3R), e.g., pramipexole, are highly effective therapy for the motor dysfunction of Parkinson's disease (PD). However, a significant subpopulation of treated individuals exhibit impulse control disorders (ICDs). To model this phenomenon, we implemented a novel probability discounting task using intracranial self-stimulation (ICSS) as the positive reinforcer and demonstrated that in PD-like rats, chronic pramipexole increases risk-taking. Emerging are the concepts that D2R/D3R activation engages the Akt/GSK-3 signaling pathway, and this cascade is involved in the trafficking and function of ionotropic glutamate receptors (AMPAR/NMDAR). The mechanisms that underlie the propensity of chronic pramipexole to dysregulate reward-motivation, especially in the parkinsonian state, are unknown, but the rich literature from the drug addiction field demonstrates that chronic treatment with indirect dopamine agonists strengthen synapses by increasing surface expression of AMPAR/NMDAR. Linking these concepts together for brain regions that regulate impulsivity, and are preserved in the parkinsonian brain state (e.g., the nucleus accumbens, NAc), will provide new insights into the cause of pramipexole-induced ICDs. Targeting this unmet need, the overall objective of this R21 project is to identify the anatomical substrates and underlying mechanisms by which pramipexole enhances impulsive behavior in a rat model of PD. As pramipexole has a high D3R/D2R affinity ratio, the central hypotheses are that chronic exposure to a therapeutically relevant dose of pramipexole enhances discounting by engaging D3R, Akt/GSK-3 pathways and AMPAR/NMDAR trafficking in the NAc, and blocking these targets will disrupt discounting. These hypotheses will be tested in two Specific Aims. For Specific Aim 1, chronic treatment with pramipexole will be used to enhance discounting, and administration of a D3R-selective antagonist, PG01037, will be used to demonstrate that D3R are necessary. We hypothesize that PG01037, given systemically or directly injected into the NAc, will block pramipexole-induced discounting. Specific Aim 2 will demonstrate that GSK-3 signaling and AMPAR/NMDAR in the NAc are necessary for pramipexole-enhanced discounting. We hypothesize that AMPAR/NMDAR surface expression will increase, and pGSK-3 (inactive) will decrease after chronic pramipexole and blockade of GSK-3 or AMPAR/NMDAR in the NAc will mitigate pramipexole- enhanced discounting. The proposed research is innovative because it reflects a novel concept regarding the adaptive consequence of chronic D3R activation in PD, and a novel model in which to validate this concept. The research is significant because it will substantially enhance our understanding of ICD neuropathology during dopamine agonist therapy for PD. Outcomes will provide a critical foundation for an R01 application for multidisciplinary, mechanistic evaluations of the neuronal consequences of such therapy, and for future development of therapeutic protocols for PD patients that provide motor benefits while avoiding ICDs.